78 EEPOET OF COMMISSIONER OF FISH AND FISHEEIES. [76] 



cup aud stem-float, in the combinatiou known as Hilgard's salinometer. 

 The water as soon as received ivS poured into bottles fitted with ground- 

 glass stoppers, and is kept in the laboratory with the instrument until 

 it has taken the tein perature of the room. The water is carefully poured 

 down the side of the cup to prevent the entanglement of air, which is 

 liable to rise in bubbles, and adheriug to the float, to vitiate the result. 

 The float is carefully inserted so as to avoid wetting the stem above the 

 line of flotation, and the graduation is read at the water level, through 

 the little cone which rises around the stem. The rolling motion of the 

 ship being very slight under all ordinary circumstances of weather, aud 

 the laboratory being situated near the center of fore and aft movement, 

 it is possible to attain here nearly the accuracy that could be expected 

 from the use of the same instruments on shore. 



Appended are the records of specific gravities (marked A). 



Particular attention is called to the interesting series of surface gravi- 

 ties taken on May ir)th and 16th, during the passage of the ship from 

 the Gulf Stream u}) Chesapeake Bay and the Potomac, and again on 

 a similar trip. The gradual and regular diminution of density and 

 the variability under differing conditions are well shown by the figures. 



Through the courtesy of the Coast Survey Office an instrument for 

 measuring the specific gravity of sea-water, known as Hilgard's opti- 

 cal densimeter, has been received. The instrument is fully described in 

 Coast Survey lieport for 1877, Appendix lS"o. 10. Much time has been 

 spent in the eftbrt to become thoroughly acquainted with the use of the 

 instrument and to determine its advanta.ges, if any, and its accuracy in 

 actual practice. After many trials, extending over a period of several 

 mouths, I am compelled to saj' that, in its present form at least, it is 

 less reliable and much more difficult to use than the old stem-float. 



The following series of trials (marked B) with distilled water illus- 

 trates the unreliability of the instrument in my hands. For each experi- 

 ment the densimeter was taken from its case, the prismatic bottle filled 

 from a large jar of distilled water, and after the examination the bottle 

 emptied and corked and the instrument returned to its case. All the 

 appliances, were kept together and were of uniform temperature at the 

 time of the exi)eriment. Every precaution was taken to make the condi- 

 tions always the same aud to secure the greatest possible accuracy in the 

 reading. As will be seen, there is an extreme variation in the series of 

 14.2 micrometer divisions, aud as each division represents a change of 

 .00007026 in density, the possible error in determining the gravity of 

 distilled water mounts up to .001. No satisfactory explanation has 

 been reached of the causes of the sudden changes observed in the 

 tables. The probable error in reading the micrometer should not ex- 

 ceed 3 divisions, and the error in reading the thermometer must be less 

 than .50. 



There is also a])peuded a table (njarked C), showing the comparative 

 results with this instrument and the salinometer, using specimens of 



